AB1133 BACK TO THE BASICS: UNDERSTANDING THE PROGNOSTIC VALUE OF RHEUMATOID FACTOR AND ANTI-CYCLIC CITRULLINATED PEPTIDE IN RHEUMATOID ARTHRITIS FROM A MUSCULOSKELETAL ULTRASOUND PERSPECTIVE

Abstract

Background:Anti-CCP and rheumatoid factor (RF), especially at high levels have been implicated as poor prognostic factors in rheumatoid arthritis (RA). With musculoskeletal ultrasound (US) being increasingly used for RA joint assessment, it is important to revisit the prognostic potential of these antibodies from an US perspective.Objectives:To study the prognostic value of RF and anti-CCP for poor disease outcomes, by investigating possible associations with US detected joint inflammation and erosion.Methods:In this cross sectional study, US power Doppler (PD) and grey scale (GS) joint inflammation were graded semi-quantitatively (0-3), while bone erosion was scored as either Yes=1 or No=0 at each joint recess. Patients disease activity was classified as either DAS28 <3.2 or DAS28 ≥3.2. Correlations of US variables with antibodies levels were quantified using the Pearson correlation coefficient. Simple linear regression was used to characterize relationships between variables and Receiver Operating Characteristic (ROC) analysis to determine statistically optimal cut-off values for identifying patients subgroups with US erosion scores >25th, >50th and >75th percentiles.Results:1080 joints and 1800 joint recesses from bilateral elbows, wrists, ankles and small joints of the hands and feet were scanned in 30 adult RA patients (76.7% Chinese and 93.3% female). Their baseline mean (SD) DAS28, disease duration, RF and anti-CCP levels were 3.58 (1.20), 70.3 (61.2) months, 142.1(179.0) IU/ml and 135.4(105.3) U/ml, respectively. Table 1 summarizes correlation coefficients of antibody levels with US variables. Among patients with DAS28 <3.2, there were no significant correlations between antibody levels with US variables. Among patients with DAS28 ≥3.2, anti-CCP levels correlated significantly (r=0.46, P=0.048) and were predictive (coefficient (95%CI), 0.028 (0.002, 0.053); R2=0.21; P=0.048) of US erosion scores. Area under the ROC curve (AUC) for anti-CCP levels (table 2) as predictive of patients with US erosion score >7 (75th percentile) was 0.72 (0.26, 0.97). For the anti-CCP level cutpoint of ≥95.2, specificity=53.8%, sensitivity=83.3%, accuracy=63.2%, Negative Predictive Value=87.5% and Positive Predictive Value=45.5%.Table 1.Correlation of RF and anti-CCP levels with ultrasound variablesDAS28<3.2UltrasoundvariablesCorrelationwith RFCorrelationwith anti-CCPCorrelationcoefficient(95% CI)P-valueCorrelationcoefficient(95% CI)P-valueErosion-0.18 (-0.71, 0.47)0.593-0.295 (-0.76, 0.37)0.379PD0.38 (-0.29, 0.80)0.249-0.287 (-0.76, 0.38)0.392GS0.381(-0.28, 0.80)0.247-0.245 (-0.74, 0.42)0.467DAS28≥3.2UltrasoundvariablesCorrelationwith RFCorrelationwith anti-CCPCorrelationcoefficient(95% CI)P-valueCorrelationcoefficient(95% CI)P-valueErosion-0.03 (-0.48, 0.43)0.9080.46 (0.01, 0.76)0.048*PD-0.26 (-0.64, 0.23)0.291-0.34 (-0.69, 0.14)0.157GS-0.01 (-0.46, 0.45)0.9750.03 (-0.44, 0.47)0.921Statistically significant: * p<0.05TABLE 2.ROC analysis of the performance of anti-CCP using various ultrasound erosion score criteriaUltrasound erosion score criterionArea under the ROC Curve (AUC) (95%CI)> 2.25 (25thpercentile)0.57 (0.26, 0.87)> 4.5 (median or 50thpercentile)0.68 (0.40, 0.95)> 7 (75thpercentile)0.72 (0.26, 0.97)11Corresponding Threshold=95.2, Specificity=53.8%, Sensitivity=83.3%, Accuracy=63.2%, NegativePredictive Value=87.5%, Positive Predictive Value=45.5%.Conclusion:The prognostic significance of anti-CCP and RF appears to differ in RA. Specifically, among patients with at least moderate disease activity (DAS28 ≥3.2), anti-CCP - but not RF - is associated with joint damage, being moderately correlated with US detected erosions.